JP5688114B2 - Method and communication apparatus in radio communication system - Google Patents

Description

  The present invention relates to a wireless communication system. In particular, the present invention relates to a network coding method in a wireless communication system.

  Network coding increases the capacity and throughput of a wireless network by mixing information received from a plurality of source nodes at a relay node and retransmitting the mixed information to one or more destination nodes. The destination node can derive the information content transmitted from the relay node based on the information entered in the relay node. By utilizing decoding techniques known in the art and combining network coding and wireless broadcast, the unicast throughput of bidirectional traffic can be increased.

  FIG. 1 shows a conventional method for exchanging information between a base station (BS) 102 and a mobile station (MS) 106 using a relay station (RS) 104 (ie, a method that does not perform network coding). In the time slot T1, the MS 106 transmits a packet “a” corresponding to the BS 102. Since the BS 102 is out of the communication range, the RS 104 obtains the packet “a” and relays it to the BS 102 in the time slot T2. In time slot T3, BS returns packet “b” to MS, which is also acquired by relay station RS and relayed at T4. Therefore, it takes time corresponding to four time slots to complete the information exchange between the BS 102 and the MS 106.

  FIG. 2 also shows how information is exchanged between the BS 102 and the MS 106. In this case, the RS 104 performs conventional network coding. In this case, the relay node (ie, RS 104) encodes and multicasts the information received from the plurality of source nodes (ie, BS 102 and MS 106). At T1, the MS 106 transmits the packet “a” to the RS 104. At T2, BS102 transmits packet “b” to RS104. At T3, the RS 104 multicasts the mixed packet “a + b” to both the BS 102 and the MS 106 (where “+” means binary XOR coding). Therefore, it takes only three time slots to complete the information exchange. The situation in FIG. 2 assumes that the single-input single-output (SISO) antenna system, BS-RS-MS system, and time slot scheduling are equal.

  Conventional wireless network coding is performed at the binary bit level in a layer higher than the network layer together with the SISO antenna system. In general, processing at a layer higher than the network layer is complicated in demodulation and decoding.

  According to existing wireless communication system techniques, one or more antennas are used at the transmitter and / or receiver. A multi-input multi-output (MIMO) wireless communication system uses a plurality of communication channels used between a plurality of antennas of a transmitter and a receiver. Therefore, in the case of a MIMO system, the transmitting device has N transmitting antennas, and the receiving device has M receiving antennas. The space-time coding method controls which data is transmitted from each of the N transmit antennas. The space-time encoding function at the transmitter processes the data to be transmitted and creates unique information for transmission from the N transmit antennas. Each of the M reception antennas receives information transmitted from each of the N transmission antennas. The space-time decoding function in the receiving apparatus combines the information transmitted from the N transmission antennas and restores the data.

  In the case of a virtual MIMO system, a plurality of mobile stations cooperate to send one mobile station data, and it looks as if it is a MIMO transmission. For example, two mobile stations having one antenna transmit one mobile station data. The two-antenna base station then receives the two signals and processes them using MIMO techniques. Adaptive virtual MIMO (Adaptive virtual MIMO) is a concept that includes a combination / mixture of virtual MIMO and non-virtual MIMO. Therefore, as a special case, there may be only virtual MIMO. Furthermore, adaptive virtual MIMO is a concept that includes virtual MIMO, single input multiple output (SIMO), or a combination of virtual MIMO and SIMO. The advantage of adaptive virtual MIMO is its flexibility to adapt to different user channel conditions.

  It is an object of the disclosed invention to provide an apparatus and method that at least alleviates the problems of the prior art.

According to one aspect of the disclosed invention,
A method of performing MIMO network coding in a wireless communication system including a relay node having a plurality of antennas, a first node, and a second node,
The first node transmitting first data to the relay node and the second node transmitting second data to the relay node;
The relay node receives transmission signals from both the first node and the second node, performs network coding on the first data and the second data using a predetermined network coding method, and is network coded. Generating information;
The relay node transmitting the network coded information to both the first node and the second node in a multi-user MIMO scheme;
The first node receives a MIMO transmission signal, performs network decoding, and restores the second data;
The second node receives a MIMO transmission signal, performs network decoding, and restores the first data.

The figure which shows the conventional method with which information exchange between BS, MS, and RS is performed without network coding. The figure which shows the method of exchanging information between BS, MS, and RS using network coding. The flowchart which shows an example which performs network coding of MIMO system. The conceptual diagram which shows the radio | wireless communication environment in the form which uses DF network coding method. Schematic which shows the mode of the bit processing between the network layer and the physical layer at the time of utilizing the DF network coding method of FIG. 4A. The table which shows the value of each variable at the time of passing through various processing steps in the form of Drawing 4A. The conceptual diagram which shows the radio | wireless communication environment in the form which uses MF network coding method. Schematic which shows the mode of the bit processing between the network layer and the physical layer at the time of utilizing the MF network coding method of FIG. 5A. The table | surface which shows the value of each variable at the time of passing through the various process steps in the form of FIG. 5A. The conceptual diagram which shows the radio | wireless communication environment in the form which uses AF network coding method. Schematic which shows the mode of the bit processing between the network layer and the physical layer at the time of utilizing the AF network coding method of FIG. 6A. The table which shows the value of each variable at the time of passing through various processing steps in the form of Drawing 6A. FIG. 4 is a flowchart illustrating a MIMO network coding architecture with details of the preliminary processing steps of FIG. 3. FIG. 4 is a flowchart illustrating a MIMO network coding architecture with details of the network coding and downlink steps of FIG. 3. FIG. 4 is a flowchart illustrating a MIMO network coding architecture with details of the network decoding step of FIG. 3. Schematic of a scheduler used in one form of the disclosed invention. The figure which shows the radio | wireless communication environment in one form. FIG. 12 is an exemplary timing diagram in the wireless communication environment of FIG. 11. 13 is a flowchart relating to the example shown in FIGS. The figure which shows the radio | wireless communication environment in one form. FIG. 15 is an exemplary timing diagram in the wireless communication environment of FIG. 16 is a flowchart relating to the example shown in FIGS. 6 is an exemplary graph showing network gain realized by MIMO network coding.

  A system and method according to one disclosed embodiment uses an adaptive virtual MIMO scheme for performing unicast transmission, network coding for the MIMO scheme, and a multi-user MIMO scheme for performing multicast transmission. The adaptive virtual MIMO scheme uses one or more mobile stations that transmit by one or more resource units.

  The network coding method used in one form is any of a decoding and forward (DF) method, a map and forward (MF) method, and an amplification and forward (AF) method.

  In one form, multi-user MIMO multicast transmission uses either space-time block coding (STC) or beamforming.

  In one form, MIMO network coding is performed at the lower physical layer, using either MF or AF coding methods.

  In one form, a simple scheduler is used to make more flexible and simple resource allocation.

  In one form, applications other than those shown in FIGS. 1 and 2 (ie, BS-RS-MS) can be used. Applications include MS-BS-MS, RS-BS-RS, MS-BS-RS, BS-RS-RS, BS-RS-MS, BS-MS-MS and RS-MS-MS. If multiple MSs are close to each other, they can form a group, which is treated as an MS group (MSG). Such situations include MSG-BS-MSG, MSG-BS-RS, BS-RS-MSG, BS-MSG-MSG and RS-MSG-MSG.

  In one form, a relay station such as RS receives an adaptive virtual MIMO transmission signal, performs MIMO network coding on the received information, and uplinks an encoded hybrid automatic repeat request (HARQ) message. To the serving station.

  In one embodiment, a MIMO network coding method is used in a wireless communication system including a relay node having a plurality of antennas, a first node, and a second node. The method includes the first node transmitting first data to the relay node, and the second node transmitting second data to the relay node, the method comprising: Receiving transmission signals from both of the two nodes, performing network coding on the first data and the second data using a predetermined network coding method, and generating network coded information. In this method, the relay node transmits network-coded information to the first node and the second node by the multi-user MIMO method, and both the first node and the second node receive the transmission signal of the MIMO method. Performing network decoding and restoring the first data and the second data.

  The first peer node belongs to a group of all peer nodes in the same coverage area. The second peer node belongs to a group of all peer nodes in the same coverage area.

  In another embodiment, a communication device (transceiver) in a wireless communication system that performs MIMO network coding is used. The communication apparatus includes a plurality of antennas and a circuit. The circuit receives the first data from the first node and the second data from the second node, and uses a predetermined network coding method. Network coding is performed on one data and second data to generate network coded information, and the network coded information is transmitted to the first node and the second node by a multi-user MIMO scheme.

  Other aspects and features of the disclosed systems and methods will become more apparent to those skilled in the art from the following description of specific forms of the disclosed invention.

  Hereinafter, the disclosed invention will be described in detail with reference to the accompanying drawings.

  In one embodiment, MIMO network coding includes the following features.

  a. Two peer nodes (or groups) transmit the relay network encoding node square method using the same or different radio resources (eg, bandwidth, time slot, etc.). Information is transmitted using one or more of spatial multiplexing, time division multiplexing, and frequency division multiplexing. In one embodiment, the information is transmitted using an adaptive virtual MIMO scheme. In this case, each of the peer node groups includes one or more nearby peer nodes. In the following description, a peer node is a concept that also includes a peer node group.

  b. The relay network encoding node receives the transmitted signal, performs network coding on the received information, and transmits the network-coded information by the multiuser MIMO scheme. In one embodiment, the MIMO transmission signal is a space division multiplexed transmission signal.

  c. Each peer node (or group) receives the MIMO stream, applies the applicable network decoding process, and restores the information.

  FIG. 3 shows an example of a flowchart for performing MIMO network coding. FIG. 3 provides a high level conceptual overview of the various steps included in each of the various forms described herein.

  FIG. 4A is a conceptual diagram showing a wireless communication environment in a form using the DF network coding method. Although all or part of the steps of FIG. 3 are applicable to the other forms described and illustrated, for the sake of simplicity of explanation, the various high-level conceptual steps of FIG. 3 are shown in FIG. 4A. The form will be described.

  The general method related to MIMO network coding shown in FIG. 3 makes it possible to perform various levels of network coding, and the various levels of network coding include binary bit level, finite field arithmetic level (finite field arithmetic level). level), modulation symbol level, signal waveform level, and the like. Various MIMO technologies including adaptive virtual MIMO, STC and beamforming can be used. This architecture can also be applied to various air interfaces including orthogonal frequency division multiplexing (OFDM), time division multiple access (TDMA), code division multiple access (CDMA), and the like.

  Scheduling step 302 relates to packet scheduling and queuing at a relay network encoding node such as BS 402 of FIG. 4A. The scheduling step 302 is executed by a scheduler, which contributes to maximizing the gain by performing MIMO network coding. Scheduling step 302 is described in further detail in connection with one form of scheduler shown in FIG.

  In step 304, preliminary processing is performed at peer nodes 404, 406. A detailed description of the preliminary processing steps will be given with reference to FIG.

  In step 306, the packet “a” is transmitted from the MS-A 404 to the BS 402 in the uplink using the virtual MIMO scheme. Further, the packet “b” is transmitted from the MS-B 406 to the BS 402 by using the virtual MIMO scheme on the uplink. The virtual MIMO uplink allows both peer nodes to transmit to the relay network encoding node using the same resource unit.

  Although not shown in FIG. 3, in practice, a demodulation step (for DF and MF) and a decoding step (for DF) are performed at BS 402 and received using a minimum mean square error (MMSE) detection technique. Decode information (or MMSE soft interference calculation (MMSE-SIC) or zero forcing method may be used).

  In step 308, network encoding is performed at the network encoding node (BS402 in FIG. 4). In this embodiment, the network coding includes binary linear combination (or finite field operation linear combination).

  In step 310, BS 402 multicasts a + b simultaneously to both MS-A 404 and MS-B 406 via a downlink MIMO transmission signal.

  FIG. 4B is a schematic diagram showing a state of bit processing between the network layer and the physical layer when the DF network coding method of FIG. 4A is used. FIG. 4C is a table showing the value of each variable when passing through various processing steps in the present embodiment.

4B and 4C are described below. Take row 1 as an example in the table of FIG. 4C. In the network layer, MS-A 404 generates information bit b _A = 0. MS- _B 406 generates information bit b _B = 0.

Both MS-A 404 and MS-B 406 pass information bits from the network layer to the physical layer, convert the information into appropriate modulation symbols, and transmit to BS 402 via resource unit # 1. For binary phase shift keying (BPSK) modulation, information bit 0 is mapped to a modulation symbol -1 (associated) Runode, the x _A = -1 and x _B = -1.

BS 402 receives the sum of two symbols from both MS-A 404 and MS-B 406. When noise is ignored, the received symbol is y _BS = −2. BS 402 uses a MIMO demodulator and forward error control (FEC) decoder to estimate the information bits (^ b _A and ^ b _B ) and pass them to the network layer. At the network layer, BS 402 handles the mixing of information and determines the network-coded bit b _BS = 0 taking into account, for example, the XOR operation. The BS takes b _BS down to the physical layer and multicasts to both MS-A 404 and MS-B 406 via resource unit # 2 using appropriate modulation (ie, x _DF = −1).

Similar considerations are possible for rows 2, 3 and 4 in the table of FIG. 4C. For row 2, MS-A 404 generates information bit b _A = 1 and MS- _B 406 generates information bit b _B = 0. For row 3, MS-A 404 generates information bit b _A = 0 and MS- _B 406 generates information bit b _B = 1. For row 4, MS-A 404 generates information bit b _A = 1 and MS- _B 406 generates information bit b _B = 1.

  A detailed description of steps 308 and 310 will be given with reference to FIG.

  In step 312, the MS-A 404 and the MS-B 406 first decode the network-coded packet, and extract desired information by performing linear operation (for example, XOR) together with the information transmitted by itself. That is, MS-A 404 decodes the transmission signal from BS 402 using the information of its own packet “a” and calculates packet “b”. Similarly, the MS-B 406 decodes the transmission signal from the BS 402 using the information of its own packet “b”, and calculates the packet “a”.

  The advantage of the example shown in FIG. 4A is that only one resource unit is used in the uplink of the virtual MIMO scheme, and only one resource unit is used in the downlink due to network coding. .

  FIG. 5A is a conceptual diagram showing a wireless communication environment in a form using the MF network coding method.

  In step 306, packet “a” is transmitted from MS-A 504 to BS 502 using the virtual MIMO scheme in the uplink. Further, the packet “b” is transmitted from the MS-B 506 to the BS 502 by using the virtual MIMO scheme on the uplink. This step can be performed simultaneously (ie, using the same resource unit).

  In step 308, network coding is performed in BS502. In the case of the present embodiment, the BS 502 maps (corresponds) the incoming symbols “a” and “b” to an appropriate symbol constellation according to the determination region. This is shown in FIG. 5B.

In step 310, BS 502 multicasts a _MF b _MF to both MS-A 504 and MS-B 506 simultaneously (eg, using STC or beamforming) using the downlink MIMO transmission signal.

  In step 312, the MS-A 504 and the MS-B 506 first decode the network-coded packet, and extract desired information by performing linear operation (for example, XOR) together with the information transmitted by the MS-A 504 and MS-B 506. That is, MS-A 504 uses the information of its own packet “a” to decode the transmission signal from BS 502 and calculate packet “b”. Similarly, the MS-B 506 uses the information of its own packet “b” to decode the transmission signal from the BS 502 and calculate the packet “a”.

  The advantage of the example shown in FIG. 5A is that only one resource unit is used in the uplink of the virtual MIMO scheme and only one resource unit is used in the downlink due to network coding. . Further, no decoding process is required at the relay node.

  FIG. 5B is a schematic diagram illustrating a state of bit processing between the network layer and the physical layer when the MF network coding method of FIG. 5A is used. FIG. 5C is a table showing the value of each variable when passing through various processing steps in the present embodiment.

Figures 5B and 5C are described below. Take row 2 as an example in the table of FIG. 5C. In the network layer, MS-A 504 generates information bit b _A = 1. MS-B506 generates information bit b _B = 0.

Both MS-A 504 and MS-B 506 pass information bits from the network layer to the physical layer, convert the information into appropriate modulation symbols, and transmit to BS 502 by resource unit # 1. In the case of BPSK modulation, information bit 1 is mapped to symbol x _A = 1 and information bit 0 is mapped to symbol x _B = −1.

BS 502 receives the sum of two symbols from both MS-A 504 and MS-B 506. If noise is ignored, the received symbol is y _BS = 0.

BS 502 uses a MIMO demodulator (no FEC decoder) and maps (corresponds) the received signal to the appropriate modulation symbol x _MF in the physical layer. An example of association rules (also referred to as mapping rules or BPSK MF decision areas) is shown on the lower side of FIG. 5B. In the case of the current example, since y _BS = 0, it is mapped to the symbol x _MF = + 1. Therefore, _xMF is a physical layer network encoded modulation symbol. Then, BS 502 multicasts the x _MF to MS-A 504 and MS-B506 both by the resource unit # 2.

  Similar considerations are possible for rows 1, 3 and 4 in the table of FIG. 5C.

  FIG. 6A is a conceptual diagram showing a wireless communication environment in a form using the AF network coding method.

  Referring to FIG. 3 again, in step 304, the peer nodes 604 and 606 perform preliminary processing.

  In step 306, the packet “a” is transmitted from the MS-A 604 to the BS 602 using the virtual MIMO scheme on the uplink. Further, the packet “b” is transmitted from the MS-B 606 to the BS 602 using the virtual MIMO scheme on the uplink. This step can be performed simultaneously (ie, using the same resource unit).

  In step 308, network coding is performed in BS602. In the case of this embodiment, the BS 602 amplifies the MIMO signal (at the waveform level). This is illustrated in FIG. 6B.

In step 310, BS 602 multicasts a _MF b _MF to both MS-A 604 and MS-B 606 simultaneously (eg, using STC or beamforming) using the downlink MIMO transmission signal.

  In step 312, MS-A 604 and MS-B 606 first subtract their own information and decode the subtracted packet to obtain desired information. That is, MS-A 604 uses the information of its own packet “a” to decode the transmission signal from BS 602 and calculate packet “b”. Similarly, the MS-B 606 uses the information of its own packet “b” to decode the transmission signal from the BS 602 and calculate the packet “a”.

  The advantage of the example shown in FIG. 6A is that only one resource unit is used in the uplink of the virtual MIMO scheme and that only one resource unit is used in the downlink due to network coding. . Furthermore, no decoding process is required at the peer node, and no demodulation process is required.

  FIG. 6B is a schematic diagram showing a state of bit processing between the network layer and the physical layer when the AF network coding method of FIG. 6A is used. FIG. 6C is a table showing the value of each variable when passing through various processing steps in the present embodiment.

6B and 6C are described below. Take the last row in the table of FIG. 6C as an example. At the network layer, MS-A 604 generates information bit b _A = 1. MS-B606 generates information bit b _B = 1.

Both MS-A 604 and MS-B 606 pass information bits from the network layer to the physical layer, convert the information into appropriate modulation symbols, and transmit to BS 602 by resource unit # 1. In the case of BPSK modulation, since information bit 1 is mapped to symbol 1, x _A = 1 and x _B = 1.

BS 602 receives the sum of two symbols from both MS-A 604 and MS-B 606. When noise is ignored, the received symbol is y _BS = 2.

_BS 602 multiplies received signal y _BS by coefficient β to obtain signal x _AF . Note that in this case, a MIMO demodulator or FEC decoder is not required.

Next, BS 602 multicasts the x _AF by the resource unit # 2 to both the MS-A 604 and MS-B 606.

  Similar considerations are possible for rows 1, 2 and 3 in the table of FIG. 6C.

  Each of FIGS. 4A, 4B, 4C, 5A, 5B, 5C, 6A, 6B, and 6C shows a specific example of a communication system or components of a communication system that can be used in embodiments of the present invention. It should be understood that the embodiments of the present invention can be implemented with a communication system having an architecture, which is different from the specific examples described herein, but with the embodiments described herein. It works so as not to contradict.

  For example, the network encoding node may be an RS node, a BS node, or an MS node, and the network decoding node may be an MS node, an RS node, or a BS node. Network coding forms are MS-BS-MS, MS-RS-MS, MS-BS-RS, RS-BS-RS, BS-RS-RS, BS-MS-MS, RS-MS-MS, MSG- Includes BS-MSG (MS group), MSG-RS-MSG, BS-MSG-MSG, RS-MSG-MSG and MSG-BS-RS. A network path (network path) can be constructed by each of the basic configurations of network coding described above or a combination thereof. A network can be constructed with each of the network coding configurations and / or network paths described above or combinations thereof. The network can use a point-to-multipoint (PMP) topology or a mesh topology.

  FIG. 7 is a flowchart showing the MIMO network coding architecture with details of the preliminary processing steps. Steps 302, 306, 308, 310 and 312 are performed in the same manner as described with reference to FIG. FIG. 7 shows only the portion related to the preliminary processing step 304 in detail.

  In one embodiment, information bit preprocessing is performed as follows. In step 702, a plurality of information bits are combined, and in step 704, a medium access control (MAC) header and a cyclic redundancy check (CRC) are added. In step 706, forward error control coding (eg, convolutional code, turbo code, low density parity check code (LDPC)) is applied. In step 708, the resulting information is mapped to modulation symbols (eg, modulation symbols are symbols in quadrature phase shift keying (QPSK), 16QAM (quadrature amplitude modulation), 64QAM). In step 710, it is determined whether the MF network coding method is used. If so, apply pre-distortion (eg, perform constellation rotation and / or power control) and go to step 714, otherwise go directly to step 714 and the baseband modulation symbol , Converted into a band-pass waveform signal. Then, the process proceeds to the uplink adaptive virtual MIMO transmission step 306 described with reference to FIG.

  FIG. 8 is a flowchart illustrating a MIMO network coding architecture with details of network coding and downlink steps. Steps 302, 304, 306 and 312 are performed in the same manner as described with reference to FIG. FIG. 8 shows in detail only the part relating to the network coding step 308 and the downlink step 310.

  In step 802, the relay node receives information / signals from both peer nodes. In step 804, it is determined which network coding method is being used. If the network coding method is MF, the process proceeds to step 806. If the network coding method is AF, the process proceeds to step 808. If the network coding method is DF, the process proceeds to step 810.

  In step 806 (ie, if the network coding method is MF), the received signal is mapped to the correct modulation symbol according to a predetermined decision area (this depends on the predistortion process in step 712). The process proceeds to step 818.

  In step 808 (ie, if the network coding method is AF), the received signal is amplified and processing proceeds to step 818.

  In step 810 (ie, if the network coding method was DF), it was post-processed using any receive processing (eg, Zend Framework (ZF), MMSE, MMSE-SIC) Find the signal. In step 812, the data stream of each peer node is demodulated, decoded and separated (de-MAC) to obtain information bits. In step 814, the network encodes information signals from both peer nodes by performing an XOR operation (or other finite field operation). If the packet sizes from both peers are different, just insert zeros for short packets before doing network coding. In step 816, a MAC header is added, FEC encoding is applied, and a modulation constellation is applied. Then, the process proceeds to Step 818.

  In step 818, the baseband signal is converted to a passband signal and the signal is multicast from the network encoding node to both peer nodes via a downlink MIMO transmission signal (eg, using STC or beamforming). The process then proceeds to the network decoding step 312 already described with respect to FIG. A more detailed description of the network decoding step will be given with reference to FIG.

  FIG. 9 is a flowchart illustrating the MIMO network coding architecture with details of the network decoding step. Steps 302, 304, 306, 308 and 310 are performed in the same manner as described with reference to FIG. FIG. 9 shows only the part relating to the network decoding step 312 in detail.

  In step 912, each peer node receives a multicast signal. In step 904, it is determined which network coding method is being used. If the network coding method is MF or DF, the process proceeds to step 906. If the network coding method is AF, the process proceeds to step 914.

  In step 906 (ie, if the network coding method is DF or MF), the signal is demodulated into the appropriate modulation symbols. In step 908, the FEC code is decoded. In step 910, the header is separated (de-MAC) (and CRC is checked) to obtain information bits (network coded). In step 912, the received (network coded) information bits and the transmitted information bits are combined by performing an XOR operation (or other finite field operation), and network decoding (decoding) is performed. Get information bits. A series of processes of demodulation 906, decoding 908, network decoding 912, and deMAC 910 may be performed iteratively. Then, the process proceeds to step 922.

  In step 914 (that is, when the network coding method is AF), the transmitted information signal is subtracted from the received multicast signal. In step 916, the subtracted signal is demodulated to an appropriate modulation symbol. In step 918, forward error control code decoding processing is performed. In step 920, the header is separated (de-MAC) (and CRC is checked) to obtain the desired information bits. Then, the process proceeds to Step 922.

  In step 922, the desired information bits are carried to the upper layer.

  FIG. 10 is a schematic diagram of a scheduler used in one form of the disclosed invention.

  A BS 1000 with a set of antennas 1004, 1006 is shown. BS 1000 is shown in wireless communication with MS-A 1008 and MS-B 1010. Note that the BS 1000 is connected to some network or part of the network, and the packet 1012 is carried to the base station 1000 via the network or the like, or the packet 1012 is carried from the base station 1000 to the network or the like. The UL and DL switch 1014 is also shown and uses this switch to send packets (in the downlink) to the BS 1000 and receive packets (in the uplink) from the BS 1000.

  A network code encoder 1016 for DF is also shown, which was received from virtual queue A1018 (ie, a queue of packets from MS-A1008) or virtual queue B1020 (ie, a queue of packets from MS-B1010) Used to encode the packet. After encoding, the packet 1012 is transferred to the UL and DL switch 1014 and transmitted to the MS-A 1008 and MS-B 1010 by the method described above.

  BS 1000 also has a MIMO network coding scheduler 1002. The MIMO network coding scheduler 1002 selects a packet to be transmitted from the packets in the virtual queue A 1018 or the virtual queue B 1020. The MIMO network coding scheduler 1002 is used to increase the network coding gain, which could not be achieved if this embodiment was not used.

  The MIMO network coding scheduler 1002 performs flexible resource allocation in the uplink and downlink. For the uplink, both peers are assigned in the same (or different) resource unit. In this case, resource units are defined as follows: (i) time-frequency subchannel in the case of OFDM, (ii) time slot in the case of TDMA, and (iii) orthogonal code in the case of CDMA. In the case of the downlink, resources are allocated for the network encoding node to perform multicast using STC and beamforming. Practical factors such as queue length and fairness are also considered (for scheduling).

  During operation, packets from MS-A 1008 (shown as solid lines) arrive at BS 1000 via UL and DL switch 1014. Packets from MS-B 1010 (shown in broken lines) arrive in a similar manner. In either case, the packet is forwarded to the MIMO network coding scheduler 1002 by the UL and DL switch 1014, and the packet enters the initial virtual queue 1022 in the order received. Then, depending on whether the packet is obtained from MS-A 1008 or MS-B 1010, the packet is transferred from the initial virtual queue 1022 to the virtual queue A 1018 or virtual queue B 1020.

  If both virtual queues are not empty, the network code encoder 1016 encodes the packets from both queues and carries it to the UL and DL switch 1014 to prepare for downlink multicast to both peers. In this way, the benefits of network coding are enhanced and system performance is enhanced. If only one queue is not empty, the network code encoder 1016 simply carries the packet from the non-empty queue to the UL and DL switch 1014. This is the same operation as that of a conventional scheduler that does not include the network code encoder 1016. If both queues are empty, the network code encoder 1016 does nothing.

  In order to increase the benefits of network coding, a policy called the lowest queue highest priority (LQHP) algorithm is used for the uplink when scheduling. This allows the scheduler 1002 to give the highest priority to users with shorter queues and increase network gain and system performance.

  An example relating to FIGS. 11 to 16 will be described. These examples are particularly suitable when using MIMO network coding related to HARQ retransmissions.

  When the relay network does not provide a network coding function at the relay station (the relay station is RS or MS, but for the sake of simplicity, RS shall mean RS or MS in the following context). ), HARQ retransmissions for various source stations (eg, mobile stations) from a relay station to a serving station (eg, BS) use different resources in the uplink direction. The example described here transfers network-coded HARQ information to a serving station using MIMO network coding. This can increase HARQ reliability and consequently reduce resource consumption.

  In one form of HARQ retransmission, the following basic steps are performed.

  (1) A plurality of MSs transmit information to a relay station using the same (or different) radio resources (for example, bandwidth, time slot). This transmission signal is partially received by the base station.

  (2) The RS receives the transmission signal, applies MIMO network coding to the received information, and transmits the encoded HARQ information signal to the serving station (for example, BS) in the uplink direction.

  (3) In addition to receiving the partial stream from the source station according to (1), the BS receives the HARQ information signal encoded according to (2) from the RS. The BS performs interleave decoding, demodulation and detection processing, and restores the original information from the MS.

  FIG. 11 is a diagram illustrating a wireless communication environment in one embodiment. A cellular radio network including a base station (BS) 1102, a relay station (RS) 1104 and two mobile stations MS1-1106, MS2-1108 is shown. BS 1102 is a serving station for MS1-1106 and MS2-1108. Being a serving station, the BS 1102 is responsible for scheduling resources for uplink transmission and HARQ retransmission (scheduling) and sending acknowledgment (ACK) / negative acknowledgment (NACK) as needed Have

  FIG. 11 shows the top geographic area of BS 1102 marked with “H” 1110. Also shown is the middle geographic area of BS 1102 marked by “M” 1112 and the lower geographic area of BS 1102 marked by “L” 1114. The coverage area of RS 1104 is marked with “C” 1116. Both MS1-1106 and MS2-1108 are located in the middle geographic area 1112 and within the coverage area 1116 of the RS 1104.

  As is well known, MSs are distributed over the coverage area of BS or RS. A first MS (eg, an MS present in a higher geographic area) that is closer to the BS than a second MS (eg, an MS present in the middle geographic area) is the BS in the uplink and downlink directions. When communicating with, it requires relatively low power.

  In the case of the current example, the synchronization and control signals of BS 1102 can reach the middle geographic region 1112. 12 is an exemplary timing diagram in the wireless communication environment of FIG. FIG. 13 is a flowchart for the example shown in FIGS.

  Referring to FIG. 11-13, in step 1302, BS 1102 performs scheduling by transmitting BS control packet B-SCH to RS 1104, MS1-1106, and MS2-1108. In step 1304, both MS1-1106 and MS2-1108 multicast their packets d1, d2 to RS 1104 and BS 1102 using the same / different resource units (in this case, by adaptive virtual MIMO scheme). .

  In step 1306, BS 1102 attempts to decode d1 and d2 received from MS1-1106 and MS2-1108, respectively. If decoding is successful due to the success / failure determination in step 1308, BS 1102 multicasts an ACK packet to MS1-1106, MS2-1108, and RS1104 in step 1316. The process ends at step 1320.

  If the decoding is not successful, BS 1102 unicasts the NACK packet to RS 1104 in step 1314. In step 1310, RS 1104 demodulates and / or decodes d1 and d2 received from MS1-1106 and MS2-1108. In step 1312, RS 1104 performs MIMO network coding (using any of the network encoding methods described above) and unicasts the encoded HARQ to BS 1102.

  In step 1306, BS 1102 collects the signals received from step 1304 and step 1312, performs repetitive network processing and channel decoding processing, and uses the original information d1, d2 transmitted from MS1-1106 and MS2-1108. get. If decoding is successful due to the success / failure determination in step 1308, BS 1102 multicasts an ACK packet to MS1-1106, MS2-1108, and RS1104 in step 1316. Otherwise, step 1314 is repeated. The process ends at step 1320.

  FIG. 14 is a diagram illustrating a wireless communication environment according to one embodiment. A cellular radio network including BS 1402, RS 1404 and two mobile stations MS1-1406, MS2-1408 is shown.

  FIG. 14 shows the geographic region above the BS 1402 marked by “H” 1410. The middle geographic area of BS 1402 marked by “M” 1412 and the lower geographic area of BS 1402 marked by “L” 1414 are also shown. The coverage area of RS1404 is marked with “C” 1416. MS1-1406 is located in the middle geographic area 1412 and MS2-1408 is located in the lower geographic area 1414. MS1-1406 and MS2-1408 are both located within the coverage area 1416 of RS1404.

  If the BS 1402 synchronization and control signals can cover the lower geographic area 1414, then the BS 1402 becomes the serving station for MS1-1406 and MS2-1408. This is called a transparent mode (transmission mode). Otherwise, RS 1404 needs to send synchronization and control signals to MS2-1408. In that case, BS 1402 is the serving station of MS1-1406, and RS1404 is the serving station of MS2-1408. This is called a non-transparent mode (non-transparent mode). The serving station (BS 1402 and possibly RS 1404) is responsible for scheduling resources for uplink transmission and HARQ retransmission (scheduling), and transmitting ACK / NACK as needed.

  In the current example, the BS 1402 synchronization and control signals reach the middle geographic region 1414 but not the lower geographic region 1414.

  FIG. 15 is an exemplary timing diagram in the wireless communication environment of FIG. FIG. 16 is a flowchart for the example shown in FIGS. In the following description, FIGS. 14-16 are referred to as appropriate.

  In step 1602, the BS 1402 performs scheduling by transmitting a BS control packet B-SCH for resource scheduling to the RS 1404 and the MS 1-1406. In step 1604, the RS 1404 performs scheduling by transmitting an RS control packet R-SCH for resource scheduling, or the BS 1402 schedules MS2-1608 (not explicitly shown but it is transparent). Mode.)

  In step 1606, MS1-1406 multicasts d1 to BS 1402 and RS1404, and MS2-1408 unicasts d2 to RS1404 using the same / different resource units.

  In step 1608, RS 1404 demodulates / decodes d1 and d2 received from MS1-1406 and MS2-1408, respectively. For convenience of explanation, this step is assumed to be successful. RS 1404 unicasts the ACK packet to MS 2-1408 (for information d2 received from MS2-1408) and unicasts the ACK packet to BS 1402 (for information d1 received from MS1-1406).

  In step 1610, BS 1402 attempts to decode d1 received from MS1-1406. If the decoding is not successful, BS 1402 unicasts the NACK packet to RS 1404, and in step 1620 RS 1404 uses MIMO network coding and d1 and d2 (using any of the above network encoding methods) and Perform both channel coding (JNCC: Joint Network and Channel Coding) (d1 and d2 are information transmitted from MS1-1406 and MS2-1408, respectively), and encoded information (JNCC-HARQ information and Unicast to BS1402.

  BS 1402 collects the signals from steps 1606 and 1620, performs iterative network processing and channel decoding processing, and obtains original information d1 from MS1-1406 and d2 from MS2-1408. If the decoding at step 1624 was successful, BS 1402 multicasts an ACK packet to MS1-1406 and RS1404, and the process ends at step 1626. If not, BS 1402 unicasts the NACK packet to RS 1404 and the process returns to step 1620.

  Referring back to step 1612, if the decoding was successful in step 1612, BS 1402 multicasts the ACK packet to MS1-1406 and RS1404. In step 1614, RS 1404 performs channel coding on d2 (from MS2-1408) and unicasts the information to BS 1402.

  BS 1402 receives the information and performs channel decoding to obtain original information d2 from MS2-1408. If the decoding was successful at step 1618, BS 1402 multicasts the ACK packet to RS 1404 and the process ends at step 1626. If not, BS 1402 unicasts the NACK packet to RS 1404 and the process returns to step 1614.

  FIG. 17 is an exemplary graph showing network gains realized by MIMO network coding. The horizontal axis indicates the BS coverage radius (in the example shown, it is 1.4 km, 1.0 km, or 0.5 km). The vertical axis shows the normalized throughput.

  The normalized throughput of six SISO systems is shown equal to 1 for comparison.

  In the case of the first system example in which the BS coverage radius is 1.4 km, the gain is improved by 8.6% by using DF network coding. In the case of the second system having a BS coverage radius of 1 km, the gain was improved by 12.99% by using DF network coding. In the case of the third system in which the BS coverage radius is 0.5 km, the gain was improved by 22.03% by using DF network coding.

  The configuration items used to examine the gain achieved by MIMO network coding are as follows. The BS uses two antennas, and the two MSs use one antenna each. Virtual MIMO is used for the uplink and space-time transmit diversity (STTD) is used for the multicast downlink with network coding. DF network coding is used. Equal throughput is planned for uplink and downlink per frame, one cell, two MSs per drop, 1000 realizations are used.

  Using the above setting items, when the BS coverage radius was 1.4 km, the MIMO gain was 88.67% and the network coding gain was 17.02%. When the BS coverage radius was 1 km, the MIMO gain was 71.06% and the network coding gain was 23.94%. When the BS coverage radius was 0.5 km, the MIMO gain was 50.87% and the network coding gain was 37.23%.

  Therefore, the overall MIMO network coding gain was seen to be greater than 85%. In this example, it was also confirmed that MIMO is purely improving the network coding gain.

  The application examples according to the principles of the present invention have been described above by way of example. Other forms and methods can be implemented by those skilled in the art without departing from the spirit and scope of the present invention.

<Related applications>
This application is filed with US Provisional Application No. 60 / 968,206, filed August 27, 2007, and US Provisional Application No. 60 / 986,682, filed with November 9, 2007. Enjoy the preferential benefits based on, and these applications are incorporated into the present application reference.

Claims (20)

A method of performing multi-input multi-output (MIMO) based network coding in a relay node in a wireless communication system including a relay node having a plurality of antennas, a first node, and a second node,
The relay node receiving first data having a first data length from the first node;
The relay node receiving second data having a second data length from the second node;
Storing the first data in a first queue and storing the second data in a second queue;
Performing network coding on at least a portion of the first data and at least a portion of the second data using a network coding method to generate network coded information;
The relay node transmitting the network coded information to the first node and the second node using a multi-user MIMO scheme;
Scheduling the transmission of packets from the first node and the second node to the relay node;
Including
A higher priority in the scheduling is given to the first node when the first queue has a shorter queue length than the second queue, and the second queue is the first queue. A method provided to the second node when having a queue length shorter than a queue.   The method according to claim 1, wherein the transmission signal of the MIMO scheme is a transmission signal of a space division multiplexing scheme.   The first data and the second data received by the relay node are transmitted to the relay node using one or more of a spatial multiplexing scheme, a time division multiplexing scheme, and a frequency division multiplexing scheme. The method of claim 1.   The method of claim 1, wherein the first node belongs to a group of nodes in the same coverage area.   The method of claim 1, wherein the second node belongs to a group of nodes in the same coverage area.   The method according to claim 1, wherein network coding is performed by performing an exclusive OR (XOR) operation on the first data and the second data at the relay node.   The method of claim 1, wherein the network coding performed at the relay node is a decoding and forward scheme (DF).   The method according to claim 1, wherein the relay node is one of a relay station (RS), a mobile station (MS), and a base station (BS). A wireless communication system including a relay node, a first node, and a second node, the relay node in a wireless communication system performing multi-input multi-output (MIMO) based network coding,
Receiving the first data having the first data length from the first node and receiving the second data having the second data length from the second node; A receiver configured to store the second data in a second queue; and
Information that is network coded by performing network coding on at least a portion of the first data from the first queue and at least a portion of the second data from the second queue using a network coding method. An encoder configured to generate
Configured to schedule transmission of packets from the first node and the second node to the relay node, wherein the first queue is shorter than the second queue with a higher priority in the scheduling A scheduler configured to assign to the first node when having a queue length and to assign to the second node when the second queue has a queue length shorter than the first queue. When,
A relay node comprising:   The relay node according to claim 9, wherein the network-coded information includes a hybrid automatic repeat request (HARQ) message. A transceiver in a wireless communication system for performing multi-input multi-output (MIMO) based network coding,
Multiple antennas,
An input configured to receive first data from a first node via the antenna and to receive second data from a second node;
A first queue configured to store the first data;
A second queue configured to store the second data;
Information that is network coded by performing network coding on at least a portion of the first data from the first queue and at least a portion of the second data from the second queue using a network coding method. A circuit configured to generate
An output configured to transmit the network coded information to the first node and the second node using a multi-user MIMO scheme;
Configured to schedule transmission of packets from the first node and the second node to the transceiver, wherein the first queue is shorter than the second queue with a higher priority in the scheduling A scheduler configured to assign to the first node when having a queue length and to assign to the second node when the second queue has a queue length shorter than the first queue. When,
Including transceiver . The transceiver according to claim 11, wherein the transmission signal of the MIMO scheme is a transmission signal by a space division multiplexing scheme. The first node transmits the first data to the transceiver using one or more of a spatial multiplexing scheme, a time division multiplexing scheme, and a frequency division multiplexing scheme, and the second node The transceiver according to claim 11, wherein the second data is transmitted to the transceiver . The transceiver according to claim 11, wherein the network coding is performed by performing an exclusive OR (XOR) operation on the first data and the second data. The transceiver according to claim 11, wherein the network coding scheme is a decoding and forward scheme (DF). The transceiver is a base station (BS), transceiver of claim 11. 12. The transceiver of claim 11, wherein the network coded information includes a hybrid automatic repeat request (HARQ) message. A network coding node in a wireless communication system for performing multi-input multi-output (MIMO) based network coding,
Multiple antennas,
An input configured to receive the first data from the first node and the second data from the second node;
A first queue configured to store the first data;
A second queue configured to store the second data;
Information that is network coded by performing network coding on at least a portion of the first data from the first queue and at least a portion of the second data from the second queue using a network coding method. A circuit configured to generate
An output configured to transmit the network coded information to the first node and the second node using a multi-user MIMO scheme;
Configured to schedule the transmission of packets from the first node and the second node to the network coding node, wherein the first queue has a higher priority than the second queue. It is configured to be assigned to the first node when it has a short queue length, and to be assigned to the second node when the second queue has a shorter queue length than the first queue. A scheduler,
Including a network coding node.   The network coding node according to claim 18 is a base station.   The relay node according to claim 9 is a base station.

Images